A method for early identification of battery system connection abnormality based on cloud operation data

Abstract

The present invention relates to the technical field of power battery fault diagnosis, specifically a method for early identification of abnormal connection of a battery system based on cloud operation data, including S1, collecting vehicle operation data from cloud storage data, extracting driving status, threshold current and SOC interval data; S2, constructing a connection abnormality risk factor, calculating the connection abnormality risk factor of the vehicle according to the vehicle operation data, the connection abnormality risk factor includes Φ1, Φ2, Φ3, Φ4, wherein Φ1, Φ2, Φ3 are respectively the first voltage related factor, the second voltage related factor, and the third voltage related factor, and Φ4 is the voltage and current comprehensive related factor; S3, establishing a connection abnormality identification rule, and using the calculation result of the connection abnormality risk factor to identify the connection abnormality. The method of the present invention can quickly identify abnormal power battery connection faults at different stages, and grade the faults to identify high-risk vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of power battery fault diagnosis, and in particular to a method for early identification of battery system connection abnormalities based on cloud operation data. Background Art

[0002] As the penetration rate of new energy vehicles is getting higher and higher, the safe and stable performance of power batteries is directly related to the safety performance of new energy vehicles. Power batteries have many types of faults, among which connection abnormalities are a typical type of abnormalities in power batteries, which are divided into two types: abnormal connections within a single cell and abnormal connections outside the cell. Connection abnormalities are mostly related to the production process of the cell. Among them, abnormal connections within the cell are mostly caused by cold welding during the welding process of the pole ear, and abnormal connections outside the cell are mostly caused by poor welding of the cell and the aluminum palladium connection, or the copper bar connection between modules. Connection abnormalities have a gradual characteristic. The characteristics in the initial stage are not obvious, and there is a problem of difficulty in identification. The long-term vibration during the use of the vehicle will cause the strength of the connection part to gradually weaken, and the contact resistance will increase accordingly, thereby increasing heat generation. Severe connection abnormalities may cause high-voltage arcing, generate high-temperature plasma, break through related components and cause thermal runaway of the battery. Therefore, it is very important to identify abnormal connection failures at an early stage.

[0003] At present, the fault diagnosis methods for abnormal connection mainly include experience-based judgment, failure mode and data-driven. Experience-based judgment and failure mode rely on manual analysis and professional knowledge of engineers or experts to diagnose faults, which have the problems of slow analysis speed and low diagnostic efficiency; data-driven uses the vehicle-side monitoring data for analysis and processing, and uses threshold alarm to complete the system fault detection. Due to the gradual characteristics of abnormal connection, it is difficult to identify it in the early stage, and the vehicle-side monitoring data cycle is short. The data is generally stored for about 7 days. Therefore, when the vehicle-side issues a fault warning, it is often at the end of the fault. At this time, the abnormal connection of the power battery has a high probability of causing abnormal power and a high risk of high-voltage arcing in the vehicle, which is not conducive to vehicle safety. At the same time, in the fault diagnosis steps, the method of analyzing a single battery cell one by one is often used, which is complex and time-consuming, and has low efficiency. Summary of the invention

[0004] The present invention provides a method for early identification of abnormal battery system connection based on cloud operation data, which can quickly identify abnormal power battery connection failures at different stages, grade the failures, and identify high-risk vehicles.

[0005] This application provides the following technical solutions:

[0006] A method for early identification of battery system connection abnormality based on cloud operation data comprises the following steps:

[0007] S1. Collecting vehicle operation data that meets the threshold current and threshold SOC range in the driving state from the cloud storage data;

[0008] S2. Construct a connection abnormality risk factor, and calculate the connection abnormality risk factor of the vehicle according to the vehicle operation data; the connection abnormality risk factor includes Φ1, Φ2, Φ3, and Φ4, wherein Φ1, Φ2, and Φ3 are respectively the first voltage-related factor, the second voltage-related factor, and the third voltage-related factor, and Φ4 is the voltage-current comprehensive related factor. The calculation process is as follows:

[0009] Extract the data under driving status, calculate the lowest voltage cell number that meets the demand current abs(I)>x, where x is the set threshold of the demand current, and the value range of x is 30~100A, calculate the proportion of each cell number, and normalize it to Φ1;

[0010] Calculate the highest voltage cell number that meets the required current abs(I)>x, calculate the proportion of each cell number, and normalize it to Φ2;

[0011] Sort the proportions, select the monomer number with a proportion greater than 90%, and use formula (1) to calculate the extreme pressure difference of the sliding SOC interval corresponding to the monomer number, which is recorded as Φ3;

[0012] Φ3=f(Volt_max, Volt_min, SOC) (1)

[0013] Formula (2) is used to calculate the ratio of the extreme pressure difference value and current in the sliding SOC interval, and the value is Φ4:

[0014] Φ4=f(I, Volt_max, Volt_min) (2)

[0015] Among them, I is the operating current, Volt_max is the maximum cell voltage of the power battery, Volt_min is the minimum cell voltage of the power battery, and SOC is the vehicle's state of charge.

[0016] S3. Establish connection anomaly identification rules and use the calculation results of the connection anomaly risk factor to identify connection anomalies.

[0017] Furthermore, the collected vehicle operation data include the highest voltage cell number, the lowest voltage cell number, the operating current, the highest cell voltage, the lowest cell voltage, the power battery SOC, and the charge and discharge status.

[0018] Furthermore, the connection anomaly identification rules include:

[0019] S31. Perform a preliminary judgment on connection abnormality based on Φ1 and Φ2 to screen vehicles with a high risk of connection abnormality;

[0020] S32. Calculate Φ3 and Φ4 for vehicles with high connection abnormality risk screened in S31;

[0021] S33. Identify the abnormal connection risk level of the power battery according to the calculation result.

[0022] Further, the S33 includes:

[0023] When Φ1 and Φ2>0.99, Φ3>0.6, and Φ4>3.5, the risk of abnormal vehicle connection is serious and the power battery needs to be disassembled and repaired immediately;

[0024] When 2.5<Φ4<3.5, the risk of abnormal vehicle connection is of secondary severity and requires intensive monitoring and timely maintenance when necessary;

[0025] When 2<Φ4<2.5, the risk of abnormal vehicle connection is at a general level and requires regular observation.

[0026] Furthermore, the method also includes step S4: conducting offline inspection and verification on vehicles with connection abnormality risks identified as severe levels.

[0027] Principles and advantages of the present invention:

[0028] The present invention proposes an early identification method for abnormal battery system connection based on cloud monitoring data, uses historical cloud data to extract fault features, uses different risk judgment factors to characterize the extraction results of fault features from multiple dimensions, and completes the determination of the abnormal connection risk of the vehicle by analyzing the judgment factors. Offline inspection and verification are carried out on vehicles identified as having a high risk of abnormal connection, and the fault is located by disassembling the power battery, verifying the accuracy of the early identification method for abnormal connection.

[0029] The method of the present invention is based on quantified typical accident data patterns, fully utilizes emergency warning detection and big data analysis methods, studies the similarities and differences in data patterns of normal vehicle and abnormal fault measurements, mines the abnormal connection risk factors of the battery system, extracts the distinguishing features of abnormally connected battery cells from different dimensions, quickly identifies abnormal power battery connection faults at different stages, and grades the faults to identify high-risk vehicles.

[0030] Compared with the vehicle-side early warning method, this method can identify and determine fault risks in advance. Limited by the vehicle-side storage configuration, the vehicle-side monitoring cycle is short, and the data is generally stored for about 7 days, so it is impossible to extract the basic data in the previous cycle for analysis. The identification method of the present invention is based on the cloud monitoring data required to be stored in accordance with the GB / 32960 Technical Specifications for Electric Vehicle Remote Service and Management System. It has a long storage cycle and complete data, which is conducive to earlier identification of connection abnormality risks.

[0031] At the same time, this method avoids the voltage calculation and analysis of each single cell, thereby improving the diagnostic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a method for early identification of battery system connection abnormality based on cloud operation data of the present invention;

[0033] Figure 2 It is a schematic diagram of the calculation process of the connection abnormality risk factor of the present invention;

[0034] Figure 3 The single cell voltage timing diagram of 6 vehicles with higher risk levels in Example 1;

[0035] Figure 4 This is a schematic diagram of the fault point of the battery of the vehicle 01 after it is disassembled in the first embodiment;

[0036] Figure 5 Schematic diagram of the fault point after the battery of vehicle 06 of embodiment 1 is disassembled. DETAILED DESCRIPTION

[0037] At present, mainstream OEMs and battery manufacturers have established cloud-based vehicle monitoring platforms, whose data fields and alarm requirements are usually stored in accordance with the technical specifications of GB / T 32960 Electric Vehicle Remote Service and Management System. The stored data fields cover the real-time operation data and static information of the vehicle, and the data cycle is more than three months. Compared with the vehicle-side monitoring method, the analysis of the longer-term data stored in the cloud-based monitoring platform can complete the identification of vehicle connection failures earlier, and carry out graded disposal according to the identification results to eliminate the safety hazards of the vehicle in a timely manner.

[0038] The following is further described in detail through specific implementation methods:

[0039] Embodiment 1

[0040] like Figure 1 As shown, a method for early identification of battery system connection abnormality based on cloud operation data includes the following steps:

[0041] S1. Collecting vehicle operation data that meets the threshold current and threshold SOC range in the driving state from the cloud storage data;

[0042] According to the storage data of the cloud monitoring platform that complies with the GB / T 32960 protocol, the standards that meet the threshold current standard and SOC range under the driving state are extracted from the database, and the vehicle operation data of the corresponding range is collected, including the highest voltage cell number, the lowest voltage cell number, the operating current, the highest cell voltage, the lowest cell voltage, the power battery SOC, and the charge and discharge status.

[0043] S2. Constructing a connection abnormality risk factor, and calculating the connection abnormality risk factor of the vehicle according to the vehicle operation data;

[0044] The connection abnormality risk factors include Φ1, Φ2, Φ3, and Φ4, where Φ1, Φ2, and Φ3 are the first voltage-related factor, the second voltage-related factor, and the third voltage-related factor, respectively, and Φ4 is a voltage-current comprehensive related factor. Figure 2 As shown, the calculation process is as follows:

[0045] Extract the data under driving status, calculate the lowest voltage cell number that meets the demand current abs(I)>x, where x is the set threshold of the demand current, and the value range of x is 30~100A, calculate the proportion of each cell number, and normalize it to Φ1;

[0046] Calculate the highest voltage cell number that meets the required current abs(I)>x, calculate the proportion of each cell number, and normalize it to Φ2;

[0047] Φ1 and Φ2 represent the proportion of the lowest voltage monomer and the highest voltage monomer that meet the threshold current standard, respectively.

[0048] The obtained quantity proportions are sorted, and the monomer number with a proportion greater than 90% is selected. The extreme pressure difference of the sliding SOC interval corresponding to the monomer number is calculated using formula (1), and the value is Φ3;

[0049] Φ3=f(Volt_max, Volt_min, SOC) (1)

[0050] Among them, Volt_max is the maximum cell voltage of the power battery, Volt_min is the minimum cell voltage of the power battery, and SOC is the vehicle's state of charge.

[0051] The step size of the SOC sliding interval is set to 100 data points. The mean of all data points in the interval is calculated and recorded as the cell voltage extreme value of the SOC sliding interval. The extreme pressure difference of the sliding interval is calculated by the extreme value of the highest cell voltage and the extreme value of the lowest cell voltage, which characterizes the degree of deviation of the cell voltage.

[0052] Formula (2) is used to calculate the ratio of the extreme pressure difference value and current in the sliding SOC interval, and the value is Φ4:

[0053] Φ4=f(I, Volt_max, Volt_min) (2)

[0054] Among them, I is the operating current, Volt_max is the maximum single cell voltage of the power battery, and Volt_min is the minimum single cell voltage of the power battery.

[0055] Φ3 and Φ4 extract the characteristics of the voltage difference extreme value and current, which can more accurately characterize the risk factor of abnormal connection failure.

[0056] S3. Establish a connection anomaly identification rule and use the calculation result of the connection anomaly risk factor to identify the connection anomaly;

[0057] The connection anomaly identification rules include:

[0058] S31. Perform a preliminary judgment on connection abnormality based on Φ1 and Φ2 to screen vehicles with a high risk of connection abnormality;

[0059] When both Φ1 and Φ2 exceed 0.9, it is preliminarily judged that there is a risk of abnormal connection, and the risk level of abnormal connection needs to be further analyzed.

[0060] S32, calculating Φ3 and Φ4 for the vehicles with high connection abnormality risk screened in S31, and performing correlation calculation on Φ1, Φ2, Φ3, and Φ4 to identify connection abnormality;

[0061] S33. Determine the abnormal connection risk level of the power battery according to the calculation result.

[0062] When Φ3>0.6 and Φ4>3.5, the risk of abnormal vehicle connection is serious and the power battery needs to be disassembled and repaired immediately;

[0063] When 2.5<Φ4<3.5, the risk of abnormal vehicle connection is of secondary severity and is judged as an early risk, which is set as key monitoring and timely maintenance when necessary;

[0064] When 2<Φ4<2.5, the risk of abnormal vehicle connection is at a general level and requires regular observation.

[0065] According to the above method, in this embodiment, the cloud data of the actual vehicle is used to check the risk of abnormal connection. The actual vehicle data comes from the defect analysis and recall effect evaluation data of the monitoring platform of the Innovation Center of the State Administration for Market Regulation. The platform has accumulated 150,000 new energy vehicles of 7 models. The platform function is mainly to monitor vehicles with market risks, check risk vehicles and evaluate the effects of vehicles involved in recalls. The monitored vehicle data includes vehicle operation data, static information, enterprise-defined alarm information, etc. The data cycle required for the algorithm operation is three months.

[0066] By checking the cloud data of more than 20,000 vehicles, 8 risky vehicles were screened out.

[0067] The vehicle power battery information is shown in Table 1: All 8 vehicles are ternary cells, pure electric models. Vehicles 01 to 03 are soft-pack cells, the modules are connected by bolts and copper bars, the cells are connected by soft aluminum-palladium, and the cells are laminated. Models 04 to 08 are square hard-shell ternary cells, the cells are connected by aluminum-palladium welding. The modules are fixed with copper bars and bolts.

[0068]

[0069]

[0070] Table 1

[0071] The calculation results of the connection abnormality risk factor are shown in Table 2:

[0072]

[0073] Table 2

[0074] As can be seen from Table 2, the higher the Φx value, the greater the deviation of the single cell voltage and the higher the risk of abnormal connection. The Φ3 value of vehicle 03 is 0.54, and the Φ4 value is 2.2. Its No. 26 single cell voltage shows a single cell outlier. It is the lowest at the discharge moment during driving, and its single cell voltage is the highest under the feedback current. The phenomenon of vehicle 05 is very similar to that of vehicle 03. By comparing vehicles 01 / 02 / 04 / 06, it can be seen that the single cell voltage deviation of the four vehicles is high, showing obvious abnormal connection phenomenon. The Φ4 values ​​of the four vehicles are all greater than 3; among them, the Φ4 of vehicle 06 is 4.2, and its single cell voltage has deviated from the median by 400mV. The Φ4 of vehicle 01 is 3.8. Judging from the calculation results, both vehicles 01 and 06 are at a serious level of abnormal connection risk.

[0075] The single cell voltage data of the six vehicles with higher risk levels were extracted, and the single cell voltage timing diagram was drawn for risk assessment. The drawing results are shown in the figure below: Figure 3 As shown in the figure, the voltage timing diagrams of the six vehicles show that the vehicles have obvious high charging and low discharging phenomenon.

[0076] S4. Conduct offline verification of vehicles identified as having serious abnormal connection risks:

[0077] Based on the recognition result of S3, vehicles 01 / 02 / 04 / 06 are selected for investigation.

[0078] in:

[0079] 1) After disassembling the power battery of vehicle 01, the fault point is as follows Figure 4 shown.

[0080] The output pole of the vehicle module is fixed with two bolts. After disassembly, it was found that the torque of the fixing bolts was only 1.7N, and the connecting copper bar was obviously loose. The contact resistance of the abnormal connection was measured using a milliohm meter. The static contact resistance of the fixed point was about 1 times higher than that of the normal connection point. After re-tightening, the vehicle returned to normal.

[0081] 2) After disassembling the power battery of vehicle 06, Figure 5 shown.

[0082] The No. 65 battery cell of the vehicle had an obvious gap in the welding area. The distance from the aluminum palladium with the battery cell to the upper cover of the battery cell was measured: No. 63 was 2.1mm, No. 64 was 2.44mm, and No. 65 was 2.7mm. The No. 65 battery cell showed obvious loose connection. As the welding area was obviously loose and the aluminum palladium could not be welded a second time, the power battery of the vehicle was scrapped.

[0083] 3) Vehicle 02 is similar to vehicle 04. After disassembly, it was found that the 46# battery cell is the module output end battery cell, and the residual torque of its output stage fixing bolt is 2.7N. It is tightened by the after-sales station and the battery packaging vehicle is resealed and put into operation.

[0084] From the inspection process and results, it can be seen that the four inspected vehicles all have process defects to varying degrees that lead to abnormal connections. The identification method of the present invention can accurately and quickly screen and identify these abnormal connection faults and determine the fault risk level.

[0085] The above are only embodiments of the present invention. The invention is not limited to the field involved in this implementation case. The common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for early identification of battery system connection abnormality based on cloud operation data, characterized in that: The following steps are involved: S1. Collecting vehicle operation data that meets the threshold current and threshold SOC range in the driving state from the cloud storage data; S2. Construct a connection abnormality risk factor. Calculate the connection abnormality risk factor of the vehicle according to the vehicle operation data. The connection abnormality risk factor includes Φ1, Φ2, Φ3, and Φ4, wherein Φ1, Φ2, and Φ3 are respectively the first voltage-related factor, the second voltage-related factor, and the third voltage-related factor, and Φ4 is the voltage-current comprehensive related factor. The calculation process is as follows: Extract the data under driving status, calculate the lowest voltage cell number that meets the demand current abs(I)>x, where x is the set threshold of the demand current, and the value range of x is 30~100A, calculate the proportion of each cell number, and normalize it to Φ1; Calculate the highest voltage cell number that meets the required current abs(I)>x, calculate the proportion of each cell number, and normalize it to Φ2; Sort the proportions, select the monomer number with a proportion greater than 90%, and use formula (1) to calculate the extreme pressure difference of the sliding SOC interval corresponding to the monomer number, which is recorded as Φ3; φ3=f(Volt_max, Volt_min, SOC) (1) Formula (2) is used to calculate the ratio of the extreme pressure difference value and current in the sliding SOC interval, and the value is Φ4: Φ4=f(I, Volt_max, Volt_min) (2) Where I is the operating current, Volt_max is the maximum cell voltage of the power battery, Volt_min is the minimum cell voltage of the power battery, and SOC is the vehicle state of charge; S3. Establish connection anomaly identification rules and use the calculation results of the connection anomaly risk factor to identify connection anomalies.

2. According to claim 1, a method for early identification of battery system connection abnormality based on cloud operation data is characterized in that: The S1 collects vehicle operation data, including the highest voltage cell number, the lowest voltage cell number, the operating current, the highest cell voltage, the lowest cell voltage, the power battery SOC, and the charge and discharge status.

3. The method for early identification of battery system connection abnormality based on cloud operation data according to claim 2 is characterized in that: The connection anomaly identification rules include: S31. Perform a preliminary judgment on connection abnormality based on Φ1 and Φ2 to screen vehicles with a high risk of connection abnormality; S32. Calculate Φ3 and Φ4 for vehicles with high connection abnormality risk screened in S31; S33. Identify the abnormal connection risk level of the power battery according to the calculation result.

4. The method for early identification of battery system connection abnormality based on cloud operation data according to claim 3 is characterized in that: The S33 includes: When Φ1 and Φ2>0.99, Φ3>0.6, and Φ4>3.5, the risk of abnormal vehicle connection is serious and the power battery needs to be disassembled and repaired immediately; When 2.5<Φ4<3.5, the risk of abnormal vehicle connection is of secondary severity and requires intensive monitoring and timely maintenance when necessary; When 2<Φ4<2.5, the risk of abnormal vehicle connection is at a general level and requires regular observation.

5. The method for early identification of battery system connection abnormality based on cloud operation data according to claim 4 is characterized in that: Also includes: S4. Conduct offline inspection and verification of vehicles with severe connection abnormality risks.

Images